Over 130,000 people in the United States live with partial or complete loss of function of the arm and hand due to injury or neurodegenerative disease. Neural prosthetic approaches could lead to therapeutic devices that have the potential to restore function and independence to this patient population. Our work is in neural prosthetic interfaces that process signals from the motor cortex to control robotic arms and hands. While much progress has been made in the neural control of a prosthetic arm, the complexities inherent in the control of the hand make the development of a neural prosthetic hand interface difficult. Current work is in the development of a neural prosthetic robot hand controller. Our aim is that it will have the flexibility to adapt to variable neural control signals while employing an automated grasp planning system to ensure stable grasping. This system will act as a bridge to complete neural control of the hand while having immediate clinical application in improving current upper limb prostheses. Specific Aim 1: To develop a robotic neural prosthetic hand control system that dynamically integrates elements of a sophisticated automated grasp planning algorithm with grasp-related information derived from neural firing in the motor cortex. Data from past human studies and current primate psychophysical experiments will be used to characterize controlled features of grasping, such as kinematic coordination of the fingers and measures of manipulability of objects with achieved grasps. Recordings from the motor cortex during grasping will be used to characterize correlation of neural firing with hand position, orientation, and shaping using population vectors, an approach used currently for the neural control of a robot arm. This data will be integrated into an existing grasp planning system to focus the search space for planning successful grasping in a way that is compatible with observed physiologic behavior and neural signal-derived intent. Specific Aim 2: To test the hypothesis that a smart hand control system can allow a primate to perform cortically controlled complete reaching and grasping tasks. The grasp planning system will be integrated into an existing setup for performing on-line neural control of an upper limb prosthetic, in which primates will perform neurally controlled robotic grasping tasks. Relation to Public Health: Neural prosthetics have the potential to help people with spinal cord injury and other neurologic disorders achieve a greater degree of independence and control of their lives. Our goal is to create a control system for a neural prosthetic hand that allows natural and effective grasping.